1324326891_94__Tamir2009

8/3/2019 1324326891_94__Tamir2009

1/6

What Do People Want to Feel and

Why?Pleasure and Utility in Emotion RegulationMaya Tamir

Boston College

ABSTRACTIt is typically assumed that people always want

to feel good. Recent evidence, however, demonstrates that

people want to feel unpleasant emotions, such as anger or

fear, when these emotions promote the attainment of theirlong-term goals. If emotions are regulatedfor instrumental

reasons, peopleshould want to feel pleasant emotions when

immediate benefits outweigh future benefits, but when fu-

ture benefits outweigh immediate benefits, people may

prefer to feel useful emotions, even if they are unpleasant.

In this article, I describe an instrumental account of emo-

tion regulation, review empirical evidence relevant to it,

and discuss its implications for promoting adaptive emo-

tional experiences.

KEYWORDSemotion regulation; self-regulation; emotion;

motivation

Research on emotion regulation has focused on how people

modify their feelings, but little research has examined why

peopledo so. People regulate their emotions to feel a certain way.

Such desired emotional states (i.e., emotional preferences) set

the course for the entire process of emotion regulation. It is

crucial, therefore, to understand what people want to feel and

why.

People want to maximize immediate pleasure. Therefore, they

want to feel pleasant emotions and avoid unpleasant ones. The

emphasis on short-term pleasure has dominated research on

emotion regulation. However, people also want to maximizeutility (i.e., long-term pleasure; Bentham, 1823/1968). There-

fore, they may also want to feel emotions that are useful (not

merely pleasurable) and avoid harmful ones. The approach that

views emotion regulation as instrumental proposes that

what people want to feel depends on both pleasure and utility

(see Fig. 1). When immediate benefits (i.e., immediate pleasure)

outweigh long-term benefits (i.e., delayed pleasure derived from

successful goal pursuit), people should prefer pleasant

emotions. When long-term benefits outweigh immediate ones,people should prefer useful emotions. The balance between

immediate and long-term benefits, in turn, depends on the goals

people pursue. This article describes the instrumental account,

reviews empirical evidence for it, and highlights its potential to

inform our understanding of emotion and its regulation.

AN INSTRUMENTAL ACCOUNT OF EMOTION

REGULATION

To identify what motivates people to regulate their emotions, the

instrumental account integrates research on self-regulation with

research on emotion. The account emphasizes that emotionregulation is a domain of self-regulation, and thus that the

principles that guide self-regulation, broadly construed, should

also guide the regulation of emotion, in particular.

Emotions involve pleasure or displeasure. Emotions can also

be useful or harmful for successful goal pursuit. By simulta-

neously recruiting multiple systems (e.g., motivation, cognition,

physiology), emotions predispose people to act in goal-condu-

cive ways (Frijda, 1986). For example, by activating the sym-

pathetic nervous system and promoting attention to threat, fear

can promote successful avoidance. In fact, fear may be more

useful for avoidance than any one specific behavior (e.g., being

vigilant), because it triggers many goal-related processes.Because emotions provide both pleasure and utility, people

may want to feel an emotion to maximize immediate pleasure,

utility, or both.

When people pursue instrumental goals (i.e., goals that secure

delayed rather than immediate reinforcement), they are willing

to forgo immediate pleasure to maximize utility (Mischel,

Shoda, & Rodriguez, 1989). For instance, although studying

hard is often unpleasant, students may want to do so when pur-

suing academic success. According to the instrumental account,

Address correspondence to Maya Tamir, Department of Psychology,

Boston College, 140 Commonwealth Ave., Chestnut Hill, MA 02467;

e-mail: [email protected].

CURRENT DIRECTIONS IN PSYCHOLOGICAL SCIENCE

Volume 18Number 2 101Copyrightr

2009 Association for Psychological Science
http://-/?-http://-/?-

8/3/2019 1324326891_94__Tamir2009

2/6

although unpleasant emotions are unpleasant to experience,

people may want to experience them when pursuing a goal that

unpleasant emotions can promote.

Wanting, in this respect, refers to seeking stimuli that pro-

mote goal attainment (i.e., that are useful) and is conceptually

and neurologically distinct from liking, which refers to seek-ing stimuli that produce immediate pleasure (Berridge & Rob-

inson, 2003). Wanting can be conscious or unconscious,

depending on the factors that give rise to it (e.g., cognitive rea-

soning or implicit learning, respectively). Thus, people may

want certain emotions whether they like them or not. What

people want to feel is not necessarily based on rational choice.

Indeed, emotional preferences can sometimes be irrationalfor

example, when one over- or underestimates the potential utility

of an emotion.

The utility of behaviors depends on the goal people pursue.

For instance, studying hard can be useful when students are

motivated to succeed academically. Therefore, when studentspursue academic success they want to study hard. The utility of

emotions likewise depends on the goal people pursue. For

instance, the experience of fear can be useful when people

are motivated to avoid threats. Therefore, from an instrumental

approach, when people are motivated to avoid threats, they

may want to feel fear and regulate their feelings accordingly.

Because the future is uncertain, preferences rely on expec-

tancies, such that people prefer stimuli they expect to be useful

(Fishbein & Ajzen, 1975). The more students motivated to

succeed academically expect studying hard to promote aca-

demic success, the more they want to study hard. Similarly,

according to the instrumental account, the more people who are

motivated to avoid threats expect fear to promote successful

avoidance, the more they should want to feel fear. The following

sections review empirical evidence in support of these predic-tions.

PEOPLE DIFFER IN WHAT THEY WANT TO FEEL

People pursue different goals and therefore should vary in the

emotions they preferto feel in particular situations. Forinstance,

individuals high in neuroticism are typically motivated to avoid

threats (Elliot & Thrash, 2002). Emotions such as fear or worry

can be useful for successful avoidance (Carver, 2001). There-

fore, individuals high in neuroticism should prefer to feel fear or

worry when anticipating possible threats. Indeed, I found that

individuals higher in neuroticism preferred to increase theirworry before taking a test (Tamir, 2005), a task they were likely

to perceive as threatening. Furthermore, feeling worried

when taking a test enhanced their performance.

Similarly, individuals high in extraversion are typically mo-

tivated to approach rewards (Elliot & Thrash, 2002). Emotions

such as happiness or excitement can be useful for successful

approach (Carver, 2001). Therefore, individuals high in extra-

version should prefer to feel happy or excited when anticipating

possible rewards. Indeed, I found that individuals higher in

Goals

Future benefits

vs.

immediate benefits

HIGH

Expected utility

of emotion

LOW HIGHHIGH

Decrease IncreaseIncrease

Pleasure in

emotion

Decrease

LOW

LOW

Fig. 1. The role of pleasure and utility in determining emotional preferences. According to the in-

strumental account, the goals people pursue affect the balance between immediate benefits (e.g.,

immediate pleasure) and long-term benefits (e.g., delayed pleasure derived from successful goal

pursuit). When immediate benefits outweigh long-term benefits, the degree to which emotions are

pleasantis a strongerdeterminantof emotional preferences,suchthat peopleshouldpreferto increase

pleasant emotions and decrease unpleasant ones. However, when long-term benefits outweigh im-

mediatebenefits, the utility of emotions is a strongerdeterminant of emotional preferences, suchthat

people should prefer to increase useful emotions and decrease harmful ones.

102 Volume 18Number 2

Pleasure and Utility in Emotion Regulation

8/3/2019 1324326891_94__Tamir2009

3/6

extraversion preferred to increase their happiness before taking

a test (Tamir, in press), a situation they were likely to perceive as

rewarding.

Other individual differences in emotional preferences have

been demonstrated, for example, as a function of age (Carsten-

sen, Fung, & Charles, 2003), self-esteem (e.g., Wood, Heimpel,

& Michela, 2003), and culture (Tsai, Miao, Seppala, Fung, &

Yeung, 2007). For instance, consistent with the instrumental

account, older adults are more likely than younger adults to seek

pleasant emotions, possibly because as one ages, immediate

benefits become more salient than long-term benefits. Such

individual differences appear to be goal related and, therefore,

maybe context dependent. Overall, there is evidence that people

differ in what they want to feel in certain contexts and that such

differences are linked to the goals they pursue.

WHEN FEELING BAD IS GOOD: GOALS AND

EMOTIONAL PREFERENCES

Different situations give rise to different goals. Therefore,emotional preferences should vary by context. Erber, Wegner,

and Therriault (1996) provided evidence for this idea, showing

that participants preferred to neutralize their feelings before

interacting with a stranger. Cases in which people want to feel

unpleasant emotions, however, provide the strongest test of the

instrumental account of emotion regulation, because in such

cases immediate pleasure diametrically contrasts with utility.

Consequently, subsequent studies tested whether people prefer

to experience unpleasant emotions when they are useful.

Anger should promote the pursuit of confrontational goals

(Frijda, 1986), so Tamir, Mitchell, and Gross (2008) tested

whether people wanted to feel angry when preparing for con-frontation. Participants were told that the study concerned

memory and computer games and that they would either recall

events or perform an unrelated activity (e.g., listen to music)

before playing. To assess emotional preferences, participants

indicated the extent to which they preferred to recall certain

events (e.g., an event in which you were angry) and listen to

musical clips preselected by the experimenter to induce either

anger, excitement, or neutral feelings before playing computer

games that were described as either confrontational (e.g., killing

enemies) or nonconfrontational (e.g., building an empire). Al-

though participants expected anger-inducing activities to be

unpleasant, they preferred to engage in them when expecting toplay the confrontational, but not the nonconfrontational, game

(see Fig. 2).

To test whether anger was useful for successful confrontation,

we manipulated participants emotional experiences by having

them listento angry, exciting, or neutral music. They then played

both a confrontational and a nonconfrontational computer game

and their performance was recorded. Angry participants per-

formed better than others in the confrontational game, by killing

more virtual enemies. They performed as well as others, how-

ever, in the nonconfrontational game. These findings demon-

strate that people may want to feel angry when anger can be

useful.

Using a similar design, Tamir and Ford (in press) tested

whether people wanted to feel afraid when preparing to avoid

threats. As predicted, when participants expected to play com-

puter games in which they had to avoid threats (e.g., escaping

monsters), they preferred to engage in fear-inducing activities

(e.g., listening to fearful music). In fact, the more participantsexpected an activity to make them afraid, the more they wanted

to engage in it before playing the threatening game.

What people want to feel before playing a computer game in

which they need to kill virtual enemies may or may not reflect

what they want to feel when they need to confront a real person.

Therefore, Tamir and Ford (2009) tested the predictions of the

instrumental approach in the context of an interpersonal nego-

tiation. Participants were told they would play the role of a

landlord and negotiate with another participant, playing the role

of a tenant who had not paid rent. Some were told that their goal

was to get their money back immediately (i.e., a confrontational

goal), some were told their goal was to maintain a long-termrelationship with the tenant (i.e., a collaboration goal), while

others were given no specific instructions (i.e., the control con-

dition). Participants then indicated what film clips they would

like to watch and what types of memories they would like to

recall before the negotiation.

As predicted, participants in the collaboration-goal condition

were more likely than others to prefer happiness-inducing ac-

tivities, whereas participants in the confrontational-goal con-

dition were more likely to prefer anger-inducing activities.

1

1.5

22.5

33.5

4

4.55

Confrontational Nonconfrontational

Context

P

reference

ratings

Angry activities Neutral activities

Exciting activities

Fig. 2. Preferences for anger-inducing, neutral, and exciting activities

when expecting to perform confrontational and nonconfrontational tasks

(Tamir, Mitchell, & Gross, 2008). People were likely to prefer anger-in-

ducing activities (i.e., listening to anger-inducing music and recalling past

events in which they were angry)when expecting to play a confrontational

computer game (i.e.,in which thethe goal was to kill virtual enemies),but

not when they expected to play a nonconfrontational game (i.e., in which

thegoal was tobuildan empire). This pattern didnotdependon thenature

of the activity that was rated (i.e., listening to music vs. recalling pastevents), suggesting that preferences were driven by the emotional tone of

the activity.

Volume 18Number 2 103

Maya Tamir

8/3/2019 1324326891_94__Tamir2009

4/6

Consistent with our predictions, anger was more useful for

confrontation, as angry participants were more likely than others

to lead their interaction partners to concede to their demands.

Taken together, these findings demonstrate that people want to

feel emotions that may be useful for attaining their goals, even

when those emotions are unpleasant.

EXPECTING FEELING BAD TO BE GOOD:

EXPECTANCIES AND EMOTIONAL PREFERENCES

If utility underlies emotional preferences, people should prefer

to feel an emotion when they expect it to be useful (Fishbein &

Ajzen, 1975). To test this hypothesis, Tamir and Ford (2009)

asked participants who prepared to negotiate with another

participant how useful they expected certain emotions to be.

As expected, preferences for anger were fully mediated by the

expected utility of anger. Participants who prepared to confront

another participant were more likely to expect anger to be useful

and, therefore, were more likely to want to feel angry before the

negotiation. Regardless of goal condition, participants who ex-pected anger to be useful wanted to be angry, whereas those who

expected happiness to be useful wanted to be happy, suggesting

that people prefer emotions they expect to be useful, whether

those emotions are actually useful or not.

MOTIVES IN EMOTION REGULATION: CONSCIOUS

VERSUS UNCONSCIOUS PROCESSES

Emotional preferences can be conscious and deliberate, yet

people may not always be aware of the factors that determine

what they want to feel. Although wanting typically refers to a

conscious desire, it can also refer to an unconscious or implicitprocess (Berridge & Robinson, 2003). Expectancies, in partic-

ular, can often operate outside of consciousness (Roese &

Sherman, 2007).

Tamir, Chiu, and Gross (2007) began to explore the link

between implicit expectancies of emotional utility and emo-

tional preferences. To assess unconscious expectancies, par-

ticipants completed a computer task in which speed of response

reflected the strength of associations between fear and utility

when avoiding threats. In the task, participants read a sentence

describing a goal (e.g., My goal is to avoid failure) and imag-

ined themselves pursuing it. Then an emotion term (e.g.,

afraid) was flashed on the screen. Finally, participants saw aword or a meaningless stringof letters andindicated whether this

was a word or not. Words in this third part of the task reflected

high or low utility (e.g., useful or harmful, respectively).

People respond faster to a word when it follows another word

that is associated with it, so we predicted that participants who

expected fear to be useful for avoiding threats would respond

faster to high-utility words than to low-utility words following

fear words in the context of avoidance goals. Participants who

showed such a pattern did not expect fear to be useful for

avoidance when asked about it explicitly. Nonetheless, they

were significantly more likely to prefer fear-inducing activities

before a threatening task. Thus, people may not necessarily

be aware of why they want to feel certain emotions in certain

situations. These findings are important, because if people are

unaware of what determines their emotional preferences,

they might have difficulty changing them.

IMPLICATIONS AND FUTURE DIRECTIONS

Research on emotion regulation has been guided by the as-

sumption that people want to feel pleasant and avoid unpleasant

emotions. The instrumental account of emotion regulation

qualifies this assumption and provides an explanatory frame-

work that gives rise to alternative testable hypotheses.

The instrumental approach offers novel accounts of functional

and dysfunctional emotion regulation. First, it highlights the

benefits of promoting knowledge about the utility of emotions. If

people know when particular emotions are useful, they may

preferand subsequently cultivateemotions that would help themattain their goals. Such knowledge should apply to pleasant and

unpleasant emotions, each promoting unique sets of goals.

Existing research examines the utility of emotions with respect

to individual goals. Future research could examine the utility

of emotions with respect to multiple, competing goals. For

instance, if collaboration promotes long-term well-being

more than competition does, happiness should be more useful

over time than anger.

Second, the instrumental approach suggests that dysfunc-

tional emotion regulation can result from wanting the wrong

emotions. This could arise from pursuing goals that are inap-

propriate in a given context (e.g., confrontational goals wheninteracting with an intimate partner) or from inaccurate expec-

tancies of emotional utility (e.g., expecting anger to be useful for

bonding with an intimate partner).Future research will test these

possibilities.

How do people know when particular emotions are useful?

Such knowledge may be acquired through learning: Learning

that an emotion is useful in one context should increase pref-

erences for that emotion in that context. One implication of the

instrumentalaccount is that it allows forthe possibility of change

in emotional preferences. An exciting avenue for future research

involves testing whether emotional preferences can be modified

through basic learning mechanisms. For instance, can a personwho prefers to feel angry to secure a partners affection learn

to prefer other feelings in that context that are less destructive?

By asking such questions, the instrumental account holds the

potential of promoting healthy emotional experiences.

Finally, much of the above research examined cases in which

people wanted to feel unpleasant emotions. However, wanting

and liking often co-occur and may be causally related (Berridge

& Robinson, 2003). If utility drives pleasure (Cabanac, 1992)

and varies by context, might the hedonic quality of emotions also

104 Volume 18Number 2

Pleasure and Utility in Emotion Regulation

8/3/2019 1324326891_94__Tamir2009

5/6

vary by context (e.g., Andrade & Cohen, 2007)? By positioning

emotion regulation in the broader realm of self-regulation, the

instrumental account promises to advance our understanding

of emotion regulation and experience.

Recommended Readings

Bonanno, G.A. (2001). Emotion self-regulation. In T.J. Mayne &G.A. Bonanno (Eds.),Emotions: Current issues and future directions

(pp. 251285). New York: Guilford. An insightful chapter that

lays the groundwork for the instrumental approach to emotion

regulation,where emotion regulationis examined within the broader

context of self-regulation.

Gross, J.J. (2007).Handbook of emotion regulation. New York:Guilford.

Provides a comprehensive review of research in the field of

emotion regulation.

Parrott, W.G. (1993). Beyond hedonism: Motives for inhibiting

good moods and for maintaining bad moods. In D.M. Wegner &

J.W. Pennebaker (Eds.), Handbook of mental control (pp. 278

305). Englewood Cliffs, NJ: Prentice Hall. The first ground-

breaking paper to argue that people may be motivated, at times, to

feel bad.Tamir, M., Mitchell, C., & Gross, J.J. (2008). Hedonic and instrumental

motives in anger regulation. Psychological Science, 19, 324328.

The first empirical demonstration that people want to feel

unpleasant emotions when those emotions are useful.

AcknowledgmentsThe author would like to thank Lisa

Feldman Barrett and Iris Mauss for their helpful comments on

a prior version of this paper.

REFERENCES

Andrade, E.B., & Cohen, J.D. (2007). On the consumption of negative

feelings. Journal of Consumer Research, 34, 283300.

Bentham, J. (1968). An introduction to the principles of morals and

legislation. In A.L. Page (Ed.), Utility Theory: A book of readings

(pp. 329). New York: Wiley. (Original work published 1823)

Berridge, K.C., & Robinson, T.E. (2003). Parsing reward. Trends in

Neurosciences, 26, 507513.

Cabanac, M. (1992). Pleasure: The common currency. Journal of

Theoretical Biology, 155, 173200.

Carstensen, L.L., Fung, H., & Charles, S. (2003). Socioemotional

selectivity theory and the regulation of emotion in the second half

of life. Motivation and Emotion, 27, 103123.

Carver, C.S. (2001). Affect and the functional bases of behavior: On the

dimensional structure of affective experience. Personality and

Social Psychology Review, 5, 345356.

Elliot, A.J., & Thrash, T.M. (2002). Approach-avoidance motivation

in personality: Approach and avoidance temperaments and goals.

Journal of Personality and Social Psychology, 82, 804818.

Erber, R., Wegner, D.M., & Therriault, N. (1996). On being cool and

collected: Mood regulation in anticipation of social interaction.


Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention, and

behavior: An introduction to theory and research. Reading, MA:

Addison-Wesley.

Frijda, N.H. (1986). The emotions. New York: Cambridge University

Press.

Mischel,W., Shoda, Y., & Rodriguez, M.L.(1989).Delay of gratification

in children. Science, 244, 933938.

Roese, N.J., & Sherman, J.W. (2007). Expectancy. In A.W. Kruglanski

& E.T. Higgins (Eds.), Handbook of social psychology: Basic

principles (2nd ed., pp. 91115). New York: Guilford.

Tamir, M. (2005). Dont worry, be happy? Neuroticism, trait-consistent

affect regulation, and performance. Journal of Personality and

Social Psychology, 89, 449461.

Tamir, M. (in press). Differential preferences for happiness: Extraver-

sion and trait-consistent emotion regulation. Journal of Person-

ality.

Tamir, M., Chiu, C.Y., & Gross, J.J. (2007). Business or pleasure?

Utilitarian versus hedonic consideration in emotion regulation.

Emotion, 7, 546554.

Tamir,M., & Ford, B. (2009). When Feeling Bad Is Expected to Be Good:

An Expectancy-Value Model of Emotion Regulation. Manuscript

submitted for publication.

Tamir, M., & Ford, B. (in press). Choosing to be afraid: Preferences for

fear as a function of goal pursuit. Emotion.

Tamir, M., Mitchell, C., & Gross, J.J. (2008). Hedonic and instrumental

motives in anger regulation. Psychological Science, 19, 324328.

Tsai, J.L., Miao, F.F., Seppala, E., Fung, H.H., & Yeung, D.Y. (2007).

Influence and adjustment goals: Sources of cultural differences in

ideal affect. Journal of Personality and Social Psychology, 92,

11021117.

Wood, J.V., Heimpel, S.A., & Michela, J.L. (2003). Savoring versus

dampening: Self-esteem differences in regulating positive affect.


Volume 18Number 2 105

Maya Tamir

8/3/2019 1324326891_94__Tamir2009

6/6

Documents

1324326891_94__Tamir2009